Home-field advantage
or a matter of ambiguity aversion?
Local bias among German individual investors
Markus Baltzer
(Deutsche Bundesbank)
Oscar Stolper
(University of Giessen)
Andreas Walter
(University of Giessen)
Discussion Paper
Series 1: Economic Studies
No 23/2011
Discussion Papers represent the authors’ personal opinions and do not necessarily reflect the views of the
Deutsche Bundesbank or its staff.
Editorial Board:
Klaus Düllmann
Frank Heid
Heinz Herrmann
Karl-Heinz Tödter
Deutsche Bundesbank, Wilhelm-Epstein-Straße 14, 60431 Frankfurt am Main,
Postfach 10 06 02, 60006 Frankfurt am Main
Tel +49 69 9566-0
Telex within Germany 41227, telex from abroad 414431
Please address all orders in writing to: Deutsche Bundesbank,
Press and Public Relations Division, at the above address or via fax +49 69 9566-3077
Internet http://www.bundesbank.de
Reproduction permitted only if source is stated.
ISBN 978-3–86558–750–3 (Printversion)
ISBN 978-3–86558–751–0 (Internetversion)
Abstract
We analyze the effect of geographic proximity on individual investors’ portfolio choice.
Using a unique data set which covers the common stockholdings of private households
at regional banks in Germany, we document strong and consistent overinvestment in
geographically close companies. Our results conclusively reject the presence of an informational advantage (‘home-field advantage’) of local over non-local investors. Instead, households’ preference for local equity turns out to be familiarity-driven. We
conclude that individual investors’ local bias is induced by ambiguity aversion in the
portfolio selection process rather than a trading strategy based on superior information
about local companies.
Keywords:
Local bias, portfolio diversification, household finance,
investor behaviour, ambiguity aversion
JEL-Classification:
G01, G11, G14
Non-technical summary
This paper investigates the role individual investors’ geographic location plays in their
equity investment decisions. Even though classic theory postulates that utilitymaximizing investors greatly benefit from holding well-diversified portfolios of risky
assets, evidence on real-life investment decisions paints a different picture. In particular,
recent research suggests that investors not only eschew foreign shares (home bias puzzle), but―in addition to this―tilt their domestic stockholdings towards locally headquartered companies. This phenomenon of disproportionately overweighting nearby
firms has been dubbed local bias in the literature and has proved robust across a variety
of countries and for private and institutional investors alike.
Yet, while the presence of local bias among investors is undisputed by now, academics still struggle to explain its causes thoroughly. Why do investors tilt their portfolios
towards local stocks? Given that local bias (a) constitutes one of retail investors’ most
fundamental deviations from what textbook models claim about optimal asset allocation
and (b) has been shown to be strong enough to move markets, finding answers to this
question is relevant for several reasons.
Several contributions to the local bias literature suggest that households’ overweight
in geographically close stocks reflects informed (i.e. rational) investment decisions
which is based on an information advantage in evaluating nearby companies (information hypothesis).
Yet, empirical evidence on informational advantages as the trigger for investors’ local bias is mixed and a variety of studies indicate that local bias, quite on the contrary, is
actually detrimental to investor welfare. If this is the case, understanding the root cause
of local bias is particularly important since it provides the basis for reducing the welfare
costs of this investment mistake. As such, a number of studies in the field soften or even
reject the information hypothesis and instead advocate that local bias is the result of
investors’ preference to invest in the familiar. However, due to the lack of a comprehensive analytical framework, these studies cannot explain exactly how investors’ familiarity with an asset actually affects local bias.
Following a theoretical concept by Boyle et al (2011), this paper investigates whether
local bias can be explained when incorporating familiarity towards stocks (and issuing
companies) as an additional dimension to the information-based portfolio selection
process. Our research is based on the Security Deposits Statistics maintained by
Deutsche Bundesbank which collects the common stockholdings of retail customers at
German regional banks on a security-by-security basis and allows specifying the geographical proximity between investor and company headquarters.
We find that, indeed, private households in Germany significantly overweight nearby
stocks and show that this result is robust across a number of different breakdowns.
Second, we investigate whether the observed portfolio locality is informationdriven―i.e. generates positive alpha―and conclusively reject the notion of a ‘homefield advantage’ for German individual investors. Finally, we test key propositions of
the framework of investor familiarity developed by Boyle et al. (2011). Our data clearly
confirms their hypotheses with regard to overinvestment in the familiar and empirically
support a flight to familiarity during financial crises. In sum, our results suggest that
including investors’ ambiguity aversion towards the available assets in the asset allocation problem contributes to explaining local bias among individual investors.
Nichttechnische Zusammenfassung
In dieser Studie untersuchen wir die Bedeutung des eigenen Standorts für die Anlageentscheidungen privater Aktieninvestoren. Obwohl die klassische Portfoliotheorie besagt, dass ein Nutzen maximierender Investor in hohem Maße von einem gut diversifizierten Portfolio riskanter Wertpapiere profitiert, zeigt sich bei der Untersuchung tatsächlich getätigter Investitionen ein anderes Bild. So offenbaren neuere Studien, dass in
den Depots privater Anleger nicht nur ausländische Aktien unterrepräsentiert sind (home bias), sondern diese darüber hinaus auch bei ihren inländischen Aktieninvestments
solche Unternehmen übergewichten, die sich im unmittelbaren Umkreis ihres Wohnorts
befinden. Dieses Phänomen der Übergewichtung lokal ansässiger Unternehmen im
Portfolio inländischer Aktienanlagen wird in der Literatur als local bias bezeichnet und
konnte empirisch für eine Vielzahl von Ländern sowie für sowohl private als auch institutionelle Investoren nachgewiesen werden. Während das Vorliegen eines solchen local
bias in der Literatur mittlerweile unbestritten ist, stellt sich weiterhin die Frage nach
einer umfassenden Erklärung für diese Verhaltensanomalie.
Einige Beiträge zur einschlägigen Literatur sehen in der Übergewichtung lokaler Aktien informierte (d.h. rationale) Anlageentscheidungen. Sie gehen von einem Informationsvorsprung bei der Bewertung lokaler Aktieninvestments aus. Allerdings ist die empirische Evidenz für Informationsvorteile als Ursache des local bias nicht eindeutig. So
zeigt eine Vielzahl anderer Studien, dass sich Anleger durch die Übergewichtung lokaler Unternehmen systematisch schlechter stellen. Dieser zweite Literaturstrang stellt die
Informationshypothese in Frage und argumentiert, dass der local bias letztlich auf eine
Präferenz des Anlegers zurückzuführen ist, in das Bekannte und Vertraute zu investieren. Aufgrund eines fehlenden umfassenden analytischen Rahmens konnten diese Studien bislang allerdings nicht klären, inwiefern die Vertrautheit eines Investors mit einer
Anlage den zu beobachtenden local bias beeinflusst.
Aufbauend auf einem theoretischen Konzept von Boyle et al. (2011) berücksichtigt
die vorliegende Studie neben dem informationsbasierten Ansatz auch die Vertrautheit
des Anlegers mit der fraglichen Aktie (bzw. dem zugrundliegenden Unternehmen) zur
Erklärung des local bias. Unsere Untersuchung basiert auf der Depotstatistik der Deutschen Bundesbank, die die Aktienbestände von Privatanlegern bei deutschen Regional-
banken erhebt und es erlaubt, die räumliche Nähe zwischen Anleger und dem Sitz der
Aktienunternehmen zu spezifizieren.
Unsere Ergebnisse belegen, dass deutsche Privathaushalte lokal ansässige Unternehmen systematisch und deutlich übergewichten. Eine umfassende Analyse der mit den
Aktien lokaler bzw. räumlich entfernter Unternehmen erwirtschafteten Renditen zeigt
außerdem, dass die These eines „Heimvorteils“ für lokale Anleger verworfen werden
muss (Informationshypothese). Schließlich prüfen wir einige der zentralen Thesen des
von Boyle et al. (2011) entwickelten Modells zur Rolle von Vertrautheit für Investoren.
Hier bestätigen unsere Ergebisse deren Hypothesen, dass Privatanleger einerseits vertraute Wertpapiere übergewichten und dass sich andererseits in Krisenphasen ein flight
to familiarity―also eine noch stärker ausgeprägte Übergewichtung in vertraute Aktien―beobachten lässt.
Zusammenfassend legen unsere Ergebnisse damit nahe, dass die Berücksichtigung
einer Vertrautheitskomponente bei der Aktienportfoliozusammensetzung einen entscheidenden Beitrag zur Erklärung des local bias bei Privatanlegern leistet.
Contents
1
Introduction and related research .............................................................................. 1
2
Data and descriptive statistics ................................................................................... 5
2.1 Data ............................................................................................................................. 5
2.2 Descriptive statistics ................................................................................................. 7
3
Do German individual investors exhibit a local equity preference? ......................... 8
3.1 Assessing the locality of investors’ stockholdings ............................................... 8
3.2 Results....................................................................................................................... 10
4
Testing the information hypothesis: Do German individual investors yield
excess returns on their local stock investments? ..................................................... 13
4.1 General intuition ...................................................................................................... 13
4.2 Methodology ............................................................................................................ 14
4.3 Results....................................................................................................................... 16
5
Testing the familiarity hypothesis: Investor ambiguity aversion and local bias ..... 19
5.1 General intuition ...................................................................................................... 19
5.2 Methodology ............................................................................................................ 20
5.3 Results....................................................................................................................... 21
6
Conclusion ............................................................................................................... 23
List of Figures
Figure 1: Geographical distribution of German individual investors and public
limited companies .......................................................................................... 29
Figure 2: Changes in local bias among German individual investors during the
sample period ................................................................................................. 30
Figure 3: Local bias of German individual investors under changing market
conditions ....................................................................................................... 31
List of Tables
Table 1: Summary statistics of sampled investor portfolios, custodian banks,
and companies ................................................................................................ 32
Table 2: Locality of German individual investors’ stockholdings ............................... 34
Table 3: Locality of German individual investors’ stockholdings, by investor
location and index status ................................................................................ 35
Table 4: Portfolio performance of German individual investors’ stockholdings
(Holdings-based portfolios, 3-month returns)................................................ 36
Table 5: Portfolio performance for investor quartiles formed on local bias levels
(Holdings-based portfolios, 3-month returns)................................................ 37
Table 6: Portfolio performance of German individual investors’ stockholdings
(Transactions-based portfolios, 12-month returns) ........................................ 38
Table 7: Impact of changes in stock market uncertainty on change in local bias
levels among German individual investors .................................................... 39
Table 8: Market performance and local bias among German individual investors ..... 40
Table 9: Variation in local bias changes across German individual investors ............. 41
Home-field advantage or a matter of ambiguity aversion?
Local bias among German individual investors∗
1
Introduction and related research
This paper investigates the role individual investors’ geographic location plays in their equity
investment decisions. Even though classic theory postulates that utility-maximizing investors
greatly benefit from holding well-diversified portfolios of risky assets, evidence on real-life
investment decisions paints a quite different picture. In particular, recent research suggests
that investors not only eschew foreign shares1, but―in addition to this―tilt their domestic
stockholdings towards locally headquartered companies. This phenomenon of disproportionately overweighting nearby firms has been dubbed local bias in the literature and has proved
robust across a variety of countries and for private and institutional investors alike.2
Yet, while the presence of local bias among investors is undisputed by now, academics still
struggle to explain its causes thoroughly. Why do investors tilt their portfolios towards local
stocks? Finding answers to this question is relevant for several reasons.
∗
Corresponding author: Markus Baltzer, Deutsche Bundesbank, Wilhelm-Epstein-Strasse 14, 60431 Frankfurt
am Main, Germany, [email protected].
Oscar Stolper, University of Giessen, Department of Financial Services, Licher Strasse 74, 35394 Giessen,
Germany, [email protected].
Andreas Walter, University of Giessen, Department of Financial Services, Licher Strasse 74, 35394 Giessen,
Germany, [email protected].
The authors would like to thank the staff of the Bundesbank Statistics Department and especially Markus
Amann and Matthias Schrape for supporting us with the relevant data. We are grateful to Ulrich Grosch and
Heinz Herrmann as well as participants of seminars at the Deutsche Bundesbank, and the University of Giessen for their helpful comments on earlier versions. The authors are exclusively responsible for any remaining
errors and inaccuracies.
1
This puzzle is referred to as home bias. See Karolyi and Stulz (2003) for a review of the home bias literature.
2
Ivkovic and Weisbenner (2005) and Seasholes and Zhu (2010) find that local stocks are overrepresented in the
equity portfolios of U.S. discount brokerage clients. Grinblatt and Keloharju (2001) provide qualitatively similar evidence for private households in Finland, Massa and Simonov (2006) and Bodnaruk (2009) document
that Swedish individual investors overweight firms with geographically close premises, and Feng and Seasholes (2004) point to a local bias among Chinese retail investors. Coval and Moskowitz (1999) show that,
while less pronounced in magnitude, local bias is also observed among U.S. fund managers.
1
On the one hand, local bias has been shown to be strong enough to move markets. Recently, Korniotis and Kumar (2009) show that stock returns feature a predictable local component.
Likewise, Hong et al. (2008) identify an ‘only-game-in-town effect’ in the presence of locally
biased investors, which characterizes a negative relation between the density of firm domiciles within a given region and the stock price levels of a company headquartered in that region. Finally, Loughran and Schultz (2004) and Jacobs and Weber (2010) show that a preference for local equity among investors also has a significant impact on firm-level turnover. In
sum, this evidence implies that identifying the reasons behind local bias may improve understanding the market impact of geography.
On the other hand, local bias is equivalent to an under-diversification of risky assets and as
such constitutes one of retail investors’ most fundamental deviations from what textbook
models claim about optimal asset allocation. Under-diversification has been identified as a
major challenge in household finance since it is assumed to have widespread effects on
household welfare.3 Interestingly enough, however, the direction in which portfolio concentration affects investors’ welfare is subject to an active debate briefly outlined in the following.
Local bias and informational advantages
Several contributions to the local bias literature suggest that households’ overweight in geographically close stocks reflects informed (i.e. rational) investment decisions. One common
approach to measure the informativeness of investment decisions is to analyze investors’ portfolio performance. The general idea is that, if investors’ preference for nearby stocks is driven
by locally generated value-relevant information, the value of that information should be reflected in an excess return of their local holdings. Related studies assume real information
asymmetries between local and remote investors and argue that information is more readily
available for local stocks. This allows local investors to form more accurate expectations
about the prospects of those stocks, thereby exploiting an information advantage in evaluating
nearby companies (information hypothesis). Indeed, several authors including Feng and Seasholes (2004), Ivkovic and Weisbenner (2005), Massa and Simonov (2006), and Bodnaruk
(2009) find that households’ local stock investments outperform their non-local ones. Note
that in these studies, local bias is not tantamount to a violation of mean-variance portfolio
3
See Campbell (2006) for a detailed discussion of this facet of household finance.
2
optimization. They argue that the increased portfolio risk incurred through the regional focus
is rewarded by a superior performance of the local stockholdings.
Local bias and a preference for the familiar
However, empirical evidence on informational advantages as the trigger for investors’ local
bias is mixed and a variety of studies indicate that local bias, quite on the contrary, is actually
detrimental to investor welfare. If this is the case, understanding the root cause of local bias is
particularly important since it provides the basis for reducing the welfare costs of this investment mistake. In a recent contribution, Seasholes and Zhu (2010) re-estimate the findings of
Ivkovic and Weisbenner (2005) using identical data and present diametrically opposed evidence of significant underperformance for U.S. households’ local equity investments. In an
earlier study, Huberman (2001) shows that shareholders of regional phone companies in the
U.S. tend to live in the area served by the company. He argues that exploiting an informational advantage essentially involves rebalancing one’s portfolio in a timely manner. Yet, his data
suggests that investors tend to buy and hold the familiar stocks, a behavior which is inconsistent with trading on information. Similarly, Grinblatt and Keloharju (2001) state, that if investors make money by exploiting information, then those investors with superior information
processing abilities should realize higher excess returns. However, in an earlier study, Grinblatt and Keloharju (2000) find that portfolio performance among Finnish investors is inversely related to investor sophistication and thus conjecture that local bias is unlikely to be driven
by information.4 Zhu (2002) finds that the local bias of retail investors decreases with growing advertisement expenditures of the companies they hold in their portfolios. He figures that
this is driven by selective attention rather than relevant information being delivered. Related
results have been obtained by Ackert et al. (2005), who indicate that local bias cannot be associated with real information asymmetries but rather with the simple fact that companies
close to home are recognizable. In an experimental analysis, they show that investors with an
otherwise identical information set perceive themselves to be more knowledgeable about
stocks in companies whose name they recognize, and subsequently overweight these securities.
All these studies soften or even reject the above-mentioned information hypothesis and instead advocate that local bias is the result of investors’ preference to invest in the familiar.
4
Barber and Odean (2000) document similar evidence for U.S. households.
3
However, due to the lack of a comprehensive analytical framework, these studies cannot exactly explain how investors’ familiarity with an asset actually affects local bias.
A comprehensive approach
This paper investigates whether local bias can be explained when incorporating familiarity as
an additional dimension to the portfolio selection process. To this end, we rely on a framework of familiarity established by Boyle et al. (2011), who build on the classic Markowitz
model but allow investors to have different degrees of ambiguity across assets. This leads to a
portfolio selection setting in which investors choose from a universe of familiar (where little
relative ambiguity pertains) and unfamiliar securities. Assuming ambiguity aversion, the
model imposes that investors optimize over risk, return, and familiarity. The resulting portfolio composition features some interesting deviations from the Markowitz-type portfolio and
offers novel, empirically testable implications. First, the optimal portfolio is biased towards
familiar assets. Second, the fraction of familiar assets increases in times of economic uncertainty, an effect which Boyle et al. (2011) call ‘flight to familiarity’.
Using geographic proximity as a proxy for familiarity towards an asset, we examine
whether this framework of familiarity is able to explain local bias among German individual
investors. Before we do so, however, we ask if German households overweight nearby stocks
at all5, and examine whether this investment behavior is nevertheless consistent with meanvariance portfolio optimization, i.e. if informational advantages may be the underlying reason
for local bias. In order to answer these questions, we study the Securities Deposits Statistics
maintained by Deutsche Bundesbank which collects the common stock investments of retail
customers at German regional banks on a security-by-security basis and allows specifying the
geographical distance between investors and company headquarters.
We find that, indeed, private households in Germany significantly overweight nearby
stocks and show that this result is robust across a number of different breakdowns. Second,
we apply comprehensive performance analysis to investigate whether the observed portfolio
locality is information-driven―i.e. generates positive alpha―and conclusively reject the notion of a ‘home-field advantage’ for German individual investors. Finally, we test key propositions of the framework of investor familiarity developed by Boyle et al. (2011). Our data
5
While not the principal objective of their work, Dorn and Huberman (2005) report that equity holdings of
clients of a German online broker are locally biased; also, the research of Hau (2001), Dorn et al. (2008), and
Jacobs and Weber (2010) points to a local equity preference among German investors. Yet, as of now, there is
no comprehensive investigation of the local bias phenomenon among German individual investors.
4
clearly confirms their hypotheses with regard to overinvestment in the familiar and empirically support a ‘flight to familiarity’ during financial crises. Taken together, our results suggest
that including investors’ ambiguity aversion towards the available assets in the asset allocation problem contributes to explaining local bias among individual investors.
The remainder of this study is organized as follows. Section 2 describes the data set. In
section 3, we introduce a measure of portfolio locality which we apply to private households’
domestic stockholdings. In section 4, we run a performance analysis to test whether the observed portfolio locality is a result of superior information about geographically close stocks.
Section 5 examines whether ambiguity aversion, on the contrary, explains investors’ local
bias. Section 6 concludes.
2
Data and descriptive statistics
2.1 Data
The database for this study is compiled from several sources. Our primary data set consists of
mandatory filings of German commercial banks for the period from December 2005 to December 2009. Each bank in Germany is required to report the aggregate quarterly shareholdings of its retail customers on a security-by-security basis. This stock data is part of a centralized register of security ownership across a variety of asset classes and investor groups maintained by the Deutsche Bundesbank for the Securities Deposits Statistics (henceforth SecuStat).6
For an investigation of investor locality, we restrict our securities sample to domestic
common stocks held by German private households at commercial banks. We confine the
universe of reported equities to shares of publicly listed companies headquartered in Germany. The resulting sample comprises 1,317 different common stocks issued by 1,109 different
corporations and effectively represents the entire universe of publicly listed companies in
Germany.
Unlike most other economies, Germany still builds upon a three-pillar commercial banking
system which consists of private banks, public savings banks and credit cooperatives. The
latter two sectors have traditionally focused on providing access to banking services for the
6
For a technical documentation of this database, see Amann et al. (2011).
5
local population.7 This distinctive feature provides us with the opportunity to geographically
demarcate their respective business spheres. We thus further narrow our sample to SecuStat
filings of savings banks and credit cooperatives. In the case of savings banks, an institution’s
outreach is typically bound to the local district it is located in. Generally, it is not possible for
savings banks to expand their activities into another institution’s business sphere.8 Analogously, cooperative banks have a mandate to promote their (local) members, and thus are also regionally bounded. Since data about the business areas of German credit cooperatives is not
available, we define a cooperative bank’s headquarter as the geographic center of its business
district.9 While we do not know the exact location of each private investor, we are reasonably
sure that customers of a certain savings or cooperative bank reside nearby the respective institution: A virtually identical portfolio of products and services within the respective banking
pillars does not provide any incentive for a customer to choose a remote institution when
there is a local one available. Consequently, they will pick the local bank for convenience,
and we assume that the holdings which a savings or cooperative bank reports, stem from local
customers. Following the approach of Grinblatt and Keloharju (2001), we define the zip code
area of the bank as the geographic center of its associated pool of investors.10 Throughout the
paper, we will refer to corresponding aggregations of private clients’ stockholdings at the
bank level as ‘(individual/private) investor’ and ‘(private) household’, respectively.11 A decentralized organizational structure of the two banking pillars together with nationwide geographic penetration leads to a high density of independent savings and cooperative banks in
the German market.12 Specifically, the resulting data set covers nearly 94% of all German
7
Wengler (2006), p. 286.
8
Until 2005, local districts typically incurred the guarantor liability of their respective regional savings bank.
Since then, banks’ geographic outreach has not changed materially.
9
This approach follows Conrad et al. (2009), p. 398.
10
Grinblatt and Keloharju (2001) use the center of the municipality in which the investor resides as the starting
point for their distance calculations.
11
Seasholes and Zhu (2010) mention that studying investor-level portfolios elevates the impact of small stock
positions and easily biases overall results; see section 4.2 for further details. In order to overcome potential
distortions, they form portfolios which aggregate the value-weighted shareholdings of many individual investors at the zip code level, which is essentially what we do.
12
Conrad et al. (2009) investigate regional variation of sector-specific bank outreach to retail customers in Germany and find that branch, deposit, and loan penetration is higher for public savings banks and credit cooperatives as compared to private banks.
6
commercial banks (1,715 out of 1,830 independent reporting entities during the period under
review).13
We match the quarterly domestic equity holdings from the SecuStat database with company-specific information on returns and free float market capitalization as well as index membership obtained from Datastream. Also, we make use of information provided by the OpenGeo Database to translate the postal codes of investors and firm headquarters into latitudinal
and longitudinal coordinates.
2.2 Descriptive statistics
Summary statistics for the sampled households and companies are reported in Table 1, while
Figure 1 plots their geographic distribution across Germany. We include the domestic shareholdings of nearly 6 million private households throughout Germany.14 Panel A of Table 1
presents basic characteristics of the average household portfolios constructed from our sample. Overall, the mean (median) value of direct investments in common stock of companies
headquartered in Germany―i.e. domestic stock―during the period under review adds up to
EUR 7,183 (EUR 6,289). Depositors living in urbanized areas of Germany account for roughly 75% of all portfolios under review and feature higher average amounts of domestic stock
investments than those in rural areas.15 Interestingly, regardless of the proximity to an urban
center, the average percentage of domestic stockholdings remains virtually identical at about
19% of households’ total portfolio value across all asset classes. Considerable heterogeneity
in the value of domestic stockholdings is however observed when comparing households in
the Western states to those in the New Länder. For the New Länder, the mean portfolio fraction held in domestic stocks declines by almost 75% to EUR 1,800 or 8.4% of average total
portfolio value. In addition, stockholders living in the New Länder constitute only 7.6% of all
portfolios under review, while the region is home to more than 16% of the German popula-
13
Note, however, that savings banks and credit cooperatives cover only roughly 36% of the total German stock
market capitalization held by domestic private households.
14
This approach differs from other local bias studies such as Ivkovic and Weisbenner (2005), Dorn and Huberman (2005), and Seasholes and Zhu (2010), among others, who infer their findings from studying the clients
of a single discount brokerage house.
15
Areas with above-median (below-median) population density are referred to as urbanized (rural). The necessary data is derived from a joint research data center run by the Federal Statistical Office and the Federal Ministry of Transport, Building, and Urban Development (INKAR), which collects respective items on an annual
basis.
7
tion.16 Taken together, however, these figures indicate that German individual investors are
less geographically concentrated than private investors in other European countries.17
Figure 1 and Table 1, Panel B, reveal that banks, as well, are much less densely distributed
in the New Länder. Only 8.5% of the sampled institutions have their premises in East Germany. This uneven spread is largely driven by the disproportionately low presence of cooperative banks, which make up nearly 75% of institutions in the full sample. Moreover, the
rightmost column of Panel B of Table 1 provides some information with regards to the banklevel aggregations of households’ portfolios employed in our subsequent analyses. On average, each bank in the sample reports the securities holdings of 3,401 private households.18
Finally, the map plotted in Figure 1 suggests that a considerable number of the firms sampled in our study cluster in only a handful of agglomeration areas, while the rest of the country is rather sparsely populated with company domiciles. Yet, with more than half of the 1,109
companies in the sample headquartered outside the ten biggest cities (Panel C of Table 1),
Germany still appears to be more evenly industrialized than other countries for which similar
empirical studies exist.19
3
Do German individual investors exhibit a local equity preference?
3.1 Assessing the locality of investors’ stockholdings
To start off, we require a distance threshold with which to classify shares that are local to a
given investor, i.e. issued by a company which is local to the investor’s home. Following the
standard approach by Coval and Moskowitz (2001), we categorize each stock within 100 kilometers of an investor’s zip code area as a local stock; shares beyond this radius are referred
to as remote or nonlocal stocks. While it can be argued that a radius of 100 kilometers is an
arbitrary threshold, we replicate our results for a number of different radii and find that they
16
As of December 31, 2009, data obtained from the Federal Statistical Office.
17
For instance, Bodnaruk (2009) reports that as much as 60% of the households analyzed in his study live in the
three largest Swedish cities. Similar proportions apply to studies conducted in Finland, cf. Grinblatt and Keloharju (2001), and Norway, cf. Doskeland and Hvide (2011).
18
Seasholes and Zhu (2010), who apply the same technique, on average aggregate the holdings of 120 households at the zip code level; see section 4.2 of this paper for further details.
19
Grinblatt and Keloharju (2001), for instance, report that as much as two thirds of all sampled firms in their
study are domiciled in the city of Helsinki.
8
do not change materially. Therefore, we stick to the radius of 100 kilometers in the following,
which makes our results more easily comparable.
In order to obtain the geographic distance between households and companies, we translate
the postal codes of each investor and each company headquarters20, respectively, into latitudes
and longitudes (measured in degrees). Using the conventional formula, we then compute the
linear distance disti, j in kilometers between investor i and stock j as:
disti , j = arccos{cos(lati ) cos(loni ) cos(lat j ) cos(lon j )
+ cos(lati ) sin(loni ) cos(lat j ) sin(lon j )
2πr
+ sin(lati ) sin(lat j )} ⋅
360
(1)
,
where lat and lon denote the latitudinal and longitudinal coordinates of the sampled investors
and companies, and r is the radius of the earth (≈6,378 kilometers). Occasionally, investor and
company headquarters share a common zip-code. In such cases, instead of assigning a zerodistance, we use one quarter of the linear distance between the pertaining zip code and the
closest neighboring postal area. This convention follows Thomas and Huggett (1980) and has
been stated customary in geographic science. Next, each stock j is assigned a weight w BM
j ,t
which corresponds to the total value of its readily available shares relative to the free float
market capitalization across the aggregate of sampled stocks at the end of the reporting period
t (last day of the respective quarter); w BM
j ,t may be interpreted as the weight of company j in the
float-adjusted market portfolio.21 Moreover, we define wi,actj,t as the actual fraction of stock j in
investor i’s portfolio at time t. We are interested in the total fraction of local investments for
each investor i. To this end, we sum up the weights wi,actj,t for all stocks j within 100 kilometers
from investor i’s place of residence at time t. Specifically, we calculate:
shi,tact =  j∈N i wi,actj,t
,
(2)
20
Ideally we would like to compute the linear distance between the investor and the closest branch or sub-sidiary
of the company considered (cf., for instance, Bodnaruk (2009) and Massa and Simonov (2006)). Yet, the necessary data is unavailable and we are encouraged by Massa and Simonov (2006, p. 652) reporting that “the
results [for either of the two alternative approaches] do not differ and the variables are highly collinear”.
21
The full market capitalization of domestic stocks also contains those assets which are not freely tradable due
to controlling shareholders and as such do not represent actual investment opportunities for individual shareholders. We use the free float market capitalization of the sampled companies to exclude the holdings of controlling shareholders when constructing our benchmark portfolios.
9
where N i denotes the number of stocks located within 100 kilometers of investor i’s place of
residence. As can be seen from section 2, however, the number of local investment opportunities varies greatly with the local area in which the investor is at home. Hence, we also compute the total fraction of available investments for each investor by summing up w BM
for all
j,t
companies headquartered within 100 kilometers of investor i’s home at time t:
shi,tBM =  j∈N i wBM
j,t
,
(3)
which is then subtracted from the fraction, the household actually invests locally. The difference between these two percentages yields our local bias metric
LBi,t = shi,tact - shi,tBM
,
(4)
This measure represents the extent to which an investor holds stocks of locally domiciled
companies in excess of what she would invest locally if she held the market portfolio.
3.2 Results
Table 2 reports empirical evidence on the degree to which an average German household’s
portfolio composition deviates from the benchmark of locally available investments. Panel A
provides an intuitive approach to assessing the local equity preference by comparing an average investor’s distance (in kilometers) from her actual portfolio versus the market portfolio
which consists of all stocks in the sample.22 The rightmost column reports the difference between the two distances and gives a first indication as to whether households actually tilt their
equity portfolios towards local companies. During the 17 quarters under review, the average
individual investor holds stocks which are 255.1 kilometers away from her place of residence,
while the distance to the market portfolio amounts to 290.5 kilometers. Hence, she invests in
stocks which are 35.4 kilometers closer than the benchmark, pointing to a substantial overweight of local companies.
Basic breakdown
Panel B of Table 2 displays the results for the local bias metric derived in section 3.1. The
average fraction of stock investments in companies headquartered within 100 kilometers of a
given household amounts to 20.1%, whereas the mean share of the market portfolio within
this radius is 11.8%. Thus, our data documents a substantial local bias of 8.3% for the period
22
For a detailed description of this measure, see Coval and Moskowitz (1999).
10
under review. This is less than the 13% excess local holdings among Norwegian households
reported by Doskeland and Hvide (2011) and the 14% local bias for U.S. retail investors documented by Seasholes and Zhu (2010).23 Interestingly, however, the inter-country difference
does not appear to stem from households’ actual holdings―at 19.6%, Seasholes and Zhu
(2010), for instance, find a virtually identical fraction of locally invested stock―but instead
from differing benchmark levels. This is intuitive, since in the U.S., the mean market capitalization within a range of 100 kilometers represents much less of the country’s aggregate market capitalization than in Germany.
Next, we check for the robustness of the basic breakdown. Specifically, we consider the
possibility that our findings are essentially the result of households residing in certain regions
and invested in certain stocks.
Local bias and investor location
It is conceivable that local bias is a phenomenon which is essentially driven by the area in
which the investor lives. Grinblatt and Keloharju (2001), for instance, provide empirical evidence that individual investors located in rural regions exhibit a particularly strong bias towards local companies. Thus, to confirm the robustness of our results with respect to investor
location, we construct subsamples according to the population structure of the household’s
local area (urbanized versus rural). Panel A of Table 3 reports the corresponding numbers and
reveals that the average local bias for residents of urbanized areas (9.7%) is even higher than
for households living in rural areas (5.8%). Consequently, we are confident that our findings
regarding the local bias extend to households living in both urbanized and rural areas.
Moreover, as discussed above in section 2.2, the geographical distribution of individual investors shows considerable variation between East and West Germany. Thus, we replicate the
analysis for the subsamples of households in the Western states and the New Länder and
show that households in the New Länder exhibit much less of a local equity preference than
those residing in the Western states (2.9% as compared to 8.8%). Yet, even for the investors
from the Eastern part of Germany, the average deviation of local stock investments from the
CAPM-efficient allocation is still highly significant. One way to rationalize the higher local
bias levels in urban versus rural areas and West Germany as compared to East Germany
23
These two studies employ the same local bias measure as we do. Qualitatively, our results are corroborated by
a number of other studies using alternative metrics, including Huberman (2001), Grinblatt and Keloharju
(2001), and Ivkovic and Weisbenner (2005), whose results also document significant local bias among individual investors.
11
might involve considering the heterogeneous geographical distribution of companies throughout Germany (see section 2.2). Company clusters in agglomeration areas provide local investors with significantly better diversification opportunities in their nearby environment than,
say, a single company in a rural East German area does. While the benchmark portfolio obviously requires a much smaller investment in local stocks in rural, less industrialized parts of
East Germany, it might not fully adjust for differences in diversification possibilities since it
accounts for company location but not for local industrial variety.
Local bias and company awareness
In addition, we would like to test whether the observed local bias is a phenomenon which is
primarily attributable to stockholdings in companies whose awareness is limited to the local
investment community. Due to higher wide-area media coverage or a greater exposure of individuals to company advertisements, for instance, some firms are visible to many potential
investors, regardless of where they reside. For a given investor, this mitigates the asymmetry
in familiarity between local and remote companies with high relative visibility. Assuming that
people tend to invest in the familiar (Huberman, 2001), this effect should reduce the fraction
of locally invested equity and thus the local bias. Following Ivkovic and Weisbenner (2005),
we choose membership in the major national stock index DAX to distinguish companies
which are nationally known from those whose awareness is likely to be regionally bounded.
DAX members are assumed to feature a relatively small potential for asymmetries in company visibility, while the opposite holds true for non-DAX companies. Panel B of Table 3 splits
the sampled equity universe into stocks of the 30 companies listed in the DAX, which together account for roughly 60% of Germany’s aggregate market capitalization, and the remainder
of stocks. Indeed, the rightmost column of Panel B reports average local bias levels of nearly
14% for the portion of non-DAX stocks, while for holdings of DAX-listed companies, the
deviation from the benchmark comes to only slightly above one third this percentage (5.3%).
This indicates that index membership partly harmonizes the differences in the awareness of a
given company between local and nonlocal investors. Alternatively, reduced local overinvestment for the subsample of DAX-listed companies might also be a result of measurement
constraints. Note that we use the sampled companies’ headquarters for our distance calculations (see footnote 20) and therefore do not capture branch-related local investments but instead count them as remote stockholdings in a world where the premises of a given company
are confined to its legal seat. Since DAX-members are particularly likely to have multiple
premises located throughout Germany, we are in turn particularly likely to underestimate lo12
cal bias levels for this subsample of firms. Again, however, the null hypothesis of no local
bias is comfortably rejected for either of the two subsamples.
Local bias and employee stock ownership
Finally, we address the question of whether our implications regarding the preference for local equity might be distorted by employee stock ownership. In Germany, large publicly traded
corporations offer employee share purchase plans (henceforth ESPP) where employees can
buy company stock at a considerable discount, if they accept a lock-up period of several
years. Assuming that most employees live in close proximity to the company they work for,
household stockholdings attributable to ESPP would appear as local investments in the data.
Our data set does not allow for a distinction of shareholders according to their affiliation to
the company they are invested in. However, this does not pose a problem, since ESPP-related
stockholdings are typically aggregated in a collective deposit held by the company for account of their employees.24 In other words, they do not appear in the SecuStat filings we examine in this study.
4
Testing the information hypothesis: Do German individual investors yield
excess returns on their local stock investments?
4.1 General intuition
In this section, we investigate whether the information hypothesis is able to explain local bias
among individual investors, i.e. whether they possess value-relevant information about local
stocks and earn abnormal returns from stock-picking. To answer this question, we analyze the
long-run performance of investors’ local stockholdings and compare it to different benchmarks. Although prior literature generally documents a poor performance of individual investors as stock market participants,25 there are several reasons why those investors might possess advantageous information for local stocks. First, since households are more exposed to
regional shocks if they tilt their portfolios towards local equity, uninformed investors should
eschew such a local overweight. Hence, the holdings in geographically close equity should
24
See Dorn and Huberman (2005, p. 469).
25
Barber (1999) and Barber and Odean (2000), for instance, document that the stockholdings of the average
broker client in their sample yield negative abnormal returns. Dorn et al. (2008) report related evidence for
German private investors.
13
reflect the investments of informed households, which is why we can expect that, on average,
local stock investments should be accompanied by excess returns if information advantages
apply. Second, by investigating nearby holdings of individual investors, we focus on the portfolio segment in which value-relevant informational asymmetries―if present―should be most
pronounced. Third and finally, we replicate the performance analysis for the subsample of
non-DAX-listed stocks in order to guarantee that our results are not weakened by the more
nationally known companies with less potential for information asymmetries between local
and nonlocal stockholders.
4.2 Methodology
We divide each investor’s portfolio into a local and nonlocal portion, using the 100 kilometer
threshold. Next, we calculate quarterly returns of the local and remote portion of her portfolio.26 These returns are then regressed on the performance of two reference portfolios.
Measuring long-run abnormal performance of individual investors
Note that several methodological issues should be considered when studying the performance
of individual investors.27 One set of pitfalls concerns the calculation of a valid test statistic.
First, we have to account for cross-sectional dependence in portfolio returns across individuals. This is necessary because our data comprises 27,819 investor-quarter observations, while
our equity universe consists of 1,317 different stocks. Thus, we have cross-correlation in returns whenever two investors hold the same stock over the same quarter. Hoechle et al. (2009)
find that test statistics which ignore cross-sectional dependence in the sample of investors’
returns can produce t-values which are three and more times higher than their correctly specified counterparts and thus are unusable. Second, we require an appropriate benchmark against
which to compare households’ returns on their local equity investments.
Moreover, empirical evidence suggests that individuals hold rather poorly diversified portfolios, typically composed of only a handful of different stocks.28 Put differently, chances are
that the local fraction of a given investor’s holdings consists of a single stock only. Thus, the
26
For details on how the returns are computed, the reader is referred to the appendix.
27
See, for instance, Lyon et al. (1999) and Hoechle et al. (2009) for problems with measuring long-run abnormal
returns of individual investor’s stockholdings as well as methodological approaches to resolve them.
28
See, for instance, Dorn and Huberman (2005) for empirical evidence of under-diversification among German
private investors.
14
monthly return of a sole stock may be counted as an observation in a standard regression
analysis, which would mean that small, volatile stocks can overly influence results.
Finally, recent empirical evidence has revealed a potential time-series selection bias when
investigating individual investors’ preference for nearby companies. As mentioned in section
1, Seasholes and Zhu (2010) re-estimate the results of Ivkovic and Weisbenner (2005) and
reach directly contradicting conclusions. They partly ascribe this to the fact that Ivkovic and
Weisbenner confine their analysis to a (arbitrarily chosen) cross-section of holdings data.
Regression specification
We make use of calendar-time portfolios in order to circumvent the problems discussed
above.29 Owing to the structure of our dataset, we build 1,715 bank-level portfolios, each of
which aggregates the stockholdings of all private households affiliated with the respective
bank, and thus ensure that the impact of small numbers of stocks is not unduly high in the
performance analysis. For each of the 1,715 portfolios, we then calculate the value-weighted
return of its local holdings. We estimate pooled ordinary least squares regressions and compute Rogers (1993) standard errors that are robust to heteroscedasticity and contemporaneous
correlation (clustered by quarter).
Also, by analyzing a time span of more than four years―with utterly different stock market
periods of boom and bust, including an unprecedented financial crisis―we implicitly avoid
arbitrary ‘snapshot’ results.
Ultimately, we compare against an additional benchmark which is specific to the location
of a given investor. This way, we address heterogeneity in the geographic distribution of industries and households, and avoid that some local investors earn superior (inferior) returns
on their local stockholdings simply because they reside in an area, where certain industries
experience higher (lower) relative returns during the period under review. Consider, for instance, the period from mid-2007 to end-2008. Those stocks in our sample, which are related
to the financial sector likely exhibit significant underperformance during that time. Recalling
the strong geographic concentration of finance-related companies in the city of Frankfurt,
chances are that we would erroneously document a significant underperformance of households in the area surrounding Frankfurt when applying the standard benchmark. To prevent
29
The calendar-time portfolio approach dates back to the work of Jaffe (1974) and Mandelker (1974) and has
proved suitable for the analysis of risk-adjusted performance of investors. See Hoechle et al. (2009) for a review of empirical finance studies applying this methodology. Our regression model largely follows Seasholes
and Zhu (2010).
15
flawed results, we therefore regress the returns of local holdings not only on a broad market
return, but also on a local benchmark. This investor-specific reference portfolio is composed
of the value-weighted market capitalization local to a given investor, i.e. stemming from
companies domiciled within 100 kilometers of her place of residence.
We form calendar-time portfolios based on both the holdings and the transactions of the
individual investors under review, and present our results in the following section.
4.3 Results
Holdings-based calendar-time portfolios
Table 4 reports the results of the holdings-based regression analysis. Regression 1 documents
the average raw excess return ( Rlocal ,i − Rf ) which households earn on their local stockholdings. In equations 2 and 3, we regress the excess local return on an overall excess market reBM
BM
turn ( Rall
- R f ) and a excess local benchmark return ( Rlocal,i
- R f ), respectively. We calculate
both benchmarks as value-weighted indices including all stocks in our sample. Note that the
market benchmark is the same for all investors, whereas the local benchmark is specific to the
geographic location of a given investor. Regression 4 represents the full specification including both the nationwide and the local benchmark. Panel A shows the results for the full stock
universe; Panel B reports the corresponding numbers for the subsample of non-DAX stocks.
We find a number of interesting results. The average quarterly excess return amounts to a
negative 1.3 basis points (bp) per quarter, which can be ascribed to the down market in the
second half of our sample period. Regressions 2 to 4 show that abnormal returns further converge to zero after adjusting for the different market betas; however, they are negative irrespective of the reference portfolio. Note that neither regression model produces significant
alphas―be it economically or statistically―, regardless of the specification we estimate. In
fact, even on an annualized basis, the gross loss before benchmark adjustment amounts to
only roughly five basis points. Returns decrease when we use the investor-specific local
benchmark as a reference portfolio (regression 3), but in terms of economic significance, results do not materially differ from those for the overall market index (regression 2).
The findings of our holdings-based analysis for the full stock universe qualitatively support
the evidence provided by Seasholes and Zhu (2010), who also document economically and
statistically insignificant alphas for U.S. individual investors (albeit positive ones). Unlike
Seasholes and Zhu (2010), however, we also investigate holdings in the subsample of non16
DAX firms (see Panel B of Table 4) and find that investors’ portfolio share of those stocks
(with presumably lower visibility for the investment community) does not generate significant
positive alpha, either. Quite on the contrary, excess returns even decrease across the board
when we replicate the performance analysis for the sample of those companies for which we
hypothesize that information asymmetries―if present―are highest. This is contrary to what
one would expect to see in case of an information-based preference for nearby equity.
Next, we test the robustness of our results by dissecting households according to how strongly
their stockholdings are tilted towards nearby companies. Assuming that investors with superior ability to pick local stocks concentrate their investments locally, whereas investors with
no such abilities hold a better diversified portfolio, it could be the case that we find abnormal
returns from nearby investments only for those households, which exhibit a high relative local
bias. To investigate this issue, we rank all households according to their investment locality
and assign them to local bias quartiles. Table 5 summarizes the results for the four resulting
portfolios. Interestingly, average local bias levels differ sharply across the quartiles. At -0.9%
on average, the 25% least locally invested households in the sample effectively show a slight
remote bias, while mean local bias levels exceed 20% for households in the top quartile. Yet,
differences across the four portfolios virtually vanish when focusing on investment performance: we re-estimate the full regression model (as described above) for each of the quartiles
and find that alphas are all indistinguishable from zero and do not differ significantly, as can
be seen in the rightmost column of Table 5. This implies that our main finding, i.e. private
households do not significantly outperform the market with their local holdings, applies to all
households in the sample, regardless of how strongly locally biased they are.
Transactions-based calendar-time portfolios
In a second step, we aim to explore if buys and sells of local stocks predict positive and negative future returns, respectively. Overall, purchases of individuals have been found to underperform their sales.30 Hence, it might be interesting to examine whether this still holds when
focusing on the portfolio fraction of geographically close stocks. To this end, we now focus
on changes in stock positions compared to the previous quarter. Note that these changes reflect net quarter-to-quarter turnover of aggregated portfolios which combine the trading decisions of many individual investors at the bank level. Certainly, this entails that opposite trades
30
See, for instance, Odean (1999), who infers this finding from analyzing the accounts of discount brokerage
customers.
17
simply cancel each other out and we capture only the lower bound of transactions. We calculate two returns for the local versus nonlocal fraction of each aggregated portfolio and distinguish between buy- and sell-positions. Hence, the dependent variables are (i) the difference in
buy
sell
- Rlocal,i
returns of an investor’s buys and sells for the local portion of her portfolio, Rlocal,i
,and―analogously―(ii) the difference in returns of the buy- and sell-positions for the portion
buy
sell
- Rremote,i
of nonlocal stocks she holds ( Rremote,i
). Returns are computed before transaction costs.
We assume that each stock is held for 12 months which has been reported as the average
holding period in related research.31
Table 6 reports regression results regarding the performance of households’ purchases and
sales of nearby stocks, at a 12-month horizon. Similar to what we see for the holdings-based
analysis, we find negative alphas: local buys underperform local sells. This time, however, the
return differential turns out to be statistically significant for both the full sample and the subsample of non-DAX companies. Interestingly, this finding again corroborates the empirical
evidence of Seasholes and Zhu (2010) for U.S. individual investors. However, they report
economically significant losses, whereas our analysis yields excess losses adding up to no
more than 5.3 bp p.a. for the full sample and 4.8 bp p.a. for the subsample of non-DAX companies. With respect to the performance of remote stocks bought minus sold, we also document a marginally negative return for the entire stock universe (-0.70 bp p.a.) as well as for
the subsample of non-DAX companies (-7.85 bp p.a.), which turns out statistically significant
for the latter group. Compared to the returns from the local portfolio fraction, this points to a
slightly better (worse) performance of the remote portfolio for the full sample (the subgroup
of non-DAX companies). However, at roughly ±3 bp p.a., this effect is marginal in magnitude
and thus economically negligible.
In sum, the results of the holdings-based as well as the transactions-based performance analysis conclusively reject the proposition that private households possess a ‘home-field advantage’ which manifests itself in value-relevant information about local companies. This has a
number of implications. First, the findings document that it is by no means rational for private
households to actively pick local stocks. In fact, returns do not compensate investors for the
concentration of diversifiable risk they hold in their portfolios when tilting them towards local
equity. Second, judging from the consistently negative return differential between local buys
31
Seasholes and Zhu (2010) document an average holding period of one year; Doskeland and Hvide (2011) state
an average holding period of 300 days.
18
and local sells it appears that, if anything, translating her local information into an investment
decision turns out to be detrimental to an individual investor’s assets.
5
Testing the familiarity hypothesis: Investor ambiguity aversion and local
bias
5.1 General intuition
So far, we have not explicitly addressed changes in local bias levels over time. In order to
examine the purely information-driven behavior by means of a performance analysis, it suffices to mitigate a potential time series selection bias, which we have been careful to do by
considering all holdings and transactions over the entire 2005 to 2009 sample period (see section 4.2). Also, this rational behavior does not provide for changes in local bias over time,
since it is unrealistic to assume that, on aggregate, investors systematically possess more information advantages at a certain point in time than before or after this date. In this section,
we investigate whether, empirically, we observe changes in individual investors’ local bias
over time and seek to rationalize them.
Our period under review is substantially different from others in that it includes extreme
market cycles. Continued GDP growth in Germany between the last quarter of 2005 and the
first quarter of 2008 is followed by four consecutive quarters of severe economic decline,
with annualized GDP plummeting by 8 percent in the last quarter of 2008 and again 14 percent in the subsequent three months. Finally, this crisis period is replaced by moderate GDP
growth from mid-2009 onwards. These heavy fluctuations are accompanied by unprecedented
stock market volatility: The broad German stock index CDAX crashes by 43% in 2008 while
in the other three years, it rises by more than 20% per annum.
We expect households to take measures in response to this strong economic downturn, i.e.
to rebalance their equity portfolios, where otherwise inertia would have prevailed.32 We are
interested in whether different business cycles have an impact on individuals’ propensity to
overinvest in nearby companies and how this teaches us new insights regarding the root cause
of local bias. Specifically, our data set allows us to test a key implication of a portfolio selection model developed by Boyle et al. (2011). The authors extend the classic Markowitz model
32
In a recent contribution, Cao et al. (2011) show analytically that investors are reluctant to trade away from
investments that they currently hold (‘status-quo bias’). In their model, a threshold has to be exceeded for a
given investor to be willing to leave this status quo.
19
by relaxing the assumption that investors are equally ambiguous about all assets. This modification is particularly suitable for asset allocation decisions, since an individual’s ambiguity
aversion is extra high in comparative situations where different chances are compared against
each other instead of being evaluated separately. They find that, when admitting investors to
exhibit different degrees of uncertainty across assets, the portfolio selection deviates from the
Markowitz theory in two important ways. The first difference concerns the portfolio composition. The optimal portfolio is now composed of a mix of familiar and unfamiliar assets. Put
differently, incorporating familiarity as a selection dimension implies that the resulting portfolio is exposed to idiosyncratic risk. The second deviation concerns the portfolio’s sensitivity
to risk. The familiarity bias triggers a rebalancing of the optimal portfolio in response to
changing asset-return correlations. In fact, Boyle et al. (2011) find that the fraction of familiar
assets increases in times of a financial crisis―an effect which they dub flight-to-familiarity.
The intuition for this result is that unfamiliar assets become less useful for diversification purposes as correlations between assets increase. An ambiguity-averse investor will therefore
hold relatively less of the unfamiliar assets. If geographic proximity is a valid proxy for familiarity, we would thus expect investors (i) to hold locally biased equity portfolios (which we
show in section 2 of this paper) and (ii) to shift their portfolio towards local stocks in times of
stock market downturn, resulting in an increased local bias during those periods. This second
implication is tested in the following.
5.2 Methodology
To test for a potential familiarity-driven investment behavior, we use expected stock return
volatility as a measure of investor uncertainty in order to gauge the negative relationship between asset price standard deviation and returns.33 To this end, we take the VDAX New index
which captures implied volatility of the DAX30 index at a one-month horizon and may be
regarded as the German equivalent to the CBOE’s market volatility index VIX based on the
S&P 500. We calculate quarterly means of the VDAX New and compare them to the corresponding local bias levels at the end of each quarter.
Before turning to the test, we are interested in the extent to which overall quarterly changes
in local bias levels actually reflect active portfolio rebalancing decisions. Note that individual
investors’ exposure to local assets may change in two ways. Obviously, they can play an ac33
See Schwert (1990), among others, for an examination of the relationship between asset price volatility and
returns.
20
tive part by trading and thereby altering the proportion of local equity in their portfolio. On
the other hand, however, they may allow price changes to naturally shift the relative weight of
nearby―as opposed to remote―companies. To distinguish these effects, we dissect the quarterly changes in local bias into a trading-induced fraction (active rebalancing) and a priceinduced fraction (passive rebalancing). Straightforwardly, the trading-driven (price-driven)
change is obtained by keeping prices (holdings) unchanged for the three-month period between two consecutive reporting dates.
Figure 2 plots the overall quarter-to-quarter change in local bias, as well as the proportions
attributable to active and passive rebalancing, respectively. Interestingly, we observe that the
transactions-based effect and the performance-based effect work in opposite directions for
nearly all quarters. In other words, including price changes understates rather than overdraws
individuals’ active decisions to shift their portfolio weights. While usually low, this attenuation is particularly pronounced around the peak of the financial crisis in the third quarter of
2008, where active rebalancing without price effects would have resulted in considerably
higher local bias levels. We conjecture that for individual investors―unlike, say, fund managers―the absence of trading does not reflect a deliberate portfolio strategy. Thus, we are
concerned with active rebalancing decisions and focus on trading-induced changes in local
bias in the following.
5.3 Results
Our findings clearly support a flight to familiarity among German individual investors. Figure
3 plots the development of the VDAX New against the respective local bias levels for the
period under review. As can be seen, we document a strong congruence between expected
stock return volatility and local bias over time. We regress the local bias on the VDAX New
to examine whether the relationship illustrated in Figure 3 proves statistically significant.
Note that, in order to do so, we use first differences of the two series since the levels turn out
to be non-stationary processes of order one. We also include an AR(1)-term to mitigate possible autocorrelation and thereby bring the Durbin-Watson statistics close towards the required
2. Finally, we alter the baseline regression (Regression 1) by adding the first lag of the VDAX
New change to capture potential inertia of households in their reaction upon a change in stock
market uncertainty (Regression 2). Table 7 reports our results. Corroborating the initial evidence from Figure 3, we find that changes in the VDAX New turn out to be highly significant
in explaining adjustments in local bias levels among private households. Specifically, a ten
21
percent rise of the VDAX New translates into an average increase in local bias of as much as
0.6 percent. Results with regard to the lag-one VDAX New change (Regression 2) are not
statistically significant, indicating that households adjust their portfolios in the same quarter
in which the change in stock market uncertainty takes place.
Interestingly, not only do we observe a flight to familiarity in times of stock market downturn, but also a reversal of this effect as soon as stock prices pick up again: when markets rebound from mid-2009 on, local bias levels also decrease significantly as a result of this development. By the end of 2009, local bias levels have almost reached pre-Lehman levels. We
further sort out the relation of market performance and local bias by calculating the quarterly
changes in German stock market performance for our period under review and assigning the
16 values we obtain into four quartiles according to magnitude of change. As a performance
measure, we use the broad CDAX index. The average change in CDAX levels per quartile,
together with the corresponding mean local bias level, is reported in Table 8. Consistent with
what we observe in Figure 3, we find that for pronounced upward or downward moving markets in particular, changes in stock market performance and local bias of German household
investors are strongly negatively correlated.
Next, we are interested in whether the portfolio rebalancing towards familiar (local) stocks
which we observe in times of economic uncertainty is the result of a few households tilting
their portfolios heavily towards nearby stocks or―on the contrary―a widespread trend
among private investors. To this end, we look at the average portfolio shift of private households aggregated at the bank level and sum up the number of mean increases and mean decreases in trading-based local bias, respectively.
Table 9 reports our results. In most quarters, we see that increases in overall local bias levels are indeed accompanied by increased average exposure to nearby companies for the
greater part of banks, and vice versa. While this relation is stronger in the second half of the
sampled period starting shortly before the financial crisis, we conclude from our data that the
flight-to-familiarity effect is driven by the majority of individual investors. Interestingly, we
see widespread shifts towards local equity following two events which private investors associate particularly strongly with the outbreak of the financial crisis. Specifically, the collapse
and firesale of Bear Stearns in mid-March 2008 was followed by a 3.4% increase in local
overinvestment, and―most prominently―, the bankruptcy of Lehman Brothers later that year
on September 15 subsequently lead to a 6.4% jump in individual investor’s local bias. Also,
Table 9 documents shifts towards geographically close companies in the second quarter of
2006, when the DAX dropped by ten percentage points in less than one month during May
22
2006. Analogously, the flight to familiarity reverses in 2009 and average increases in local
equity reach their lowest levels in mid-2009 when newsflow has again become positive across
the board and several important economic indicators reach their pre-Lehman levels.
Taken together, we find (i) that individual investors pull out of remote (unfamiliar) stocks,
and pour into local (familiar) stocks during times of financial crises, (ii) that this flight to familiarity is driven by active portfolio rebalancing and rebounds when the economy picks up
again, and, finally, (iii) that this shift is a robust phenomenon across the majority of individual
investors.
Given this evidence, there is reason to assume that the concept of investor familiarity induced by ambiguity aversion might be a promising avenue in the attempt to explain the local
bias phenomenon among individual investors. One can reasonably assume that private households do not devote their entire time and energy to collecting and processing all information
available in the market. Put differently, due to time (and other) constraints, it is unrealistic to
conjecture that people are equally ambiguous about all securities. Rather, they possess different degrees of ambiguity across different securities, as modeled by Boyle et al. (2011), and
favor investments which they are less ambiguous about. In addition to that, portfolio choices
are characterized by several distinct features which may aggravate investors’ ambiguity aversion. On the one hand, asset allocation decisions are situations in which different chances are
compared against each other instead of being evaluated separately. In such settings, people are
particularly inclined to exhibit intolerance towards the uncertain (Fox and Tversky, 1995). On
the other hand, investment choices are decisions where the majority of households judge
themselves to be relatively less competent; this matches the findings of Heath and
Tversky (1991), who demonstrate that ambiguity aversion is especially pronounced when
people find it difficult to assess a set of prospects. Consequently, local bias seems to be part
of a larger phenomenon in which individual investors show a preference for the familiar
which is attributable to ambiguity aversion. Investors’ aversion to risk implies that their portfolios should be diversified, but―being ambiguity averse―they trade off a piece of this diversification by overweighting familiar assets in their stock portfolios, thereby creating a local
bias.
6
Conclusion
This paper contributes to the literature on the geography of investment. We investigate the
role of individual investors’ location for their stockholdings. Analyzing a rich data set which
23
covers the accounts of nearly six million private households in Germany, we find strong evidence for a local bias―i.e. an overweight of geographically close versus remote German
companies in their equity portfolios―and investigate the reasons for this portfolio concentration. We consider two possible explanations. On the one hand, it could be that the overweight
in nearby stocks reflects informed investment decisions. In this case, a locally biased portfolio
would not necessarily constitute an investment mistake. Instead, it could still be consistent
with traditional mean-variance portfolio theory in case the increased portfolio risk incurred
through the regional focus is rewarded by a superior performance of the local stockholdings.
The intuition is that, if investors’ preference for geographically close companies is driven by
locally generated value-relevant information, then the value of that information should be
reflected in an excess return of their local stocks. More readily available information allows
investors to form more refined expectations about those stocks, thereby exploiting an information advantage when assessing these nearby companies (information hypothesis). In order
to test for potential information advantages, we conduct a comprehensive analysis of both the
holdings-based and the transactions-based portfolio performance of private households in our
sample. Our results conclusively reject a ‘home-field advantage’ of local over non-local individual investors. Next, we investigate to what extent investors’ familiarity with nearby companies is able to explain the empirically observed local bias. While prior studies in the field
suggest that familiarity might be the reason for local bias, they remain largely unclear about
the nature and the influence of familiarity on portfolio concentration. We take a theoretical
framework of investor familiarity developed by Boyle et al. (2011) to the data. The model
claims that familiarity arises from investors’ aversion towards the ambiguous and predicts that
investors are not only locally biased but―in addition to that―local bias should increase in
times of economic uncertainty (‘flight to familiarity’ effect). The latter proposition is inconsistent with an information-based explanation. Using geographic proximity as a proxy for
familiarity, we find clear support for investor familiarity as the underlying reason of locally
biased equity portfolios. We conclude that individual investors’ local bias is a matter of ambiguity aversion in the portfolio selection process rather than a trading strategy based on locally
generated information advantages. The finding that, on average, private households in Germany do not benefit from a concentration of local assets in their portfolios, however, suggests
that they are well-advised not to confuse familiarity with information about a company when
making investment decisions.
24
References
Ackert, Lucy F., Bryan K. Church, James Tompkins, and Ping Zhang (2005) What's in a
Name? An Experimental Examination of Investment Behavior, Review of Finance 9 (2),
281–304.
Amann, Markus, Markus Baltzer, and Matthias Schrape (2011) Microdatabase: Securities
Depostis Statistics. A flexible multi-dimensional approach for providing user-targeted
securities holdings data, Deutsche Bundesbank, Mimeo.
Barber, Brad M., and Terrance Odean (2000) Trading Is Hazardous to Your Wealth: The
Common Stock Investment Performance of Individual Investors, Journal of Finance 55
(2), 773–806.
Benartzi, Shlomo (2001) Excessive Extrapolation and the Allocation of 401(k) Accounts to
Company Stock, Journal of Finance 56 (5), 1747–1764.
Bodnaruk, Andriy (2009) Proximity Always Matters: Evidence from Swedish Data, Review of
Finance 13 (4), 629–656.
Boyle, Phelim, Lorenzo Garlappi, Raman Uppal, and Tan Wang (2011) Keynes Meets Markowitz: The Trade-off Between Familiarity and Diversification, Management Science,
forthcoming.
Cao, H. Henry, Bing Han, David Hirshleifer, and Harold H. Zhang (2011) Fear of the Unknown: Familiarity and Economic Decisions, Review of Finance 15 (1), 173–206.
Conrad, Alexander, Doris Neuberger, and Maria Schneider-Reißig (2009) Geographic and
Demographic Bank Outreach: Evidence from Germany’s Three-Pillar Banking System,
Kredit und Kapital 42 (3), 377–411.
Coval, Joshua D., and Tobias J. Moskowitz (1999) Home Bias at Home: Local Equity Preference in Domestic Portfolios, Journal of Finance 54 (6), 2045–2073.
Coval, Joshua D., and Tobias J. Moskowitz (2001) The Geography of Investment: Informed
Trading and Asset Prices, Journal of Political Economy 109 (4), 811–841.
Deutsche Bank Monthly Report March (2003).
Dorn, Daniel, and Gur Huberman (2005) Talk and Action: What Investors Say and What
They Do, Review of Finance 9 (4), 437–481.
Dorn, Daniel, Gur Huberman, and Paul Sengmueller (2008) Correlated Trading and Returns,
Journal of Finance 63 (2), 885–920.
Doskeland, Trond M., and Hans K. Hvide (2011) Do Individual Investors Have Asymmetric
Information Based on Work Experience?, Journal of Finance 66 (3), 1011–1041.
Fox, Craig R., and Amos Tversky (1995) Ambiguity Aversion and Cognitive Ignorance,
Quarterly Journal of Economics 110 (3), 585–603.
Goetzman, William N., and Alok Kumar (2008) Equity Portfolio Diversification, Review of
Finance 12 (3), 433–463.
Grinblatt, Mark, and Matti Keloharju (2000) The Investment Behavior and Performance of
Various Investor Types – A Study of Finland’s Unique Data Set, Journal of Financial
Economics 55 (1), 43–67.
25
Grinblatt, Mark, and Matti Keloharju (2001) How Distance, Language, and Culture Influence
Stockholdings and Trades, Journal of Finance 56 (3), 1053–1073.
Heath, Chip, and Amos Tversky (1991) Preference and Belief – Ambiguity and Competence
in Choice under Uncertainty, Journal of Risk and Uncertainty 4 (1), 5–28.
Hau, Harald (2001) Location Matters: An Examination of Trading Profits, Journal of Finance
56 (5), 1959–1983.
Hoechle, Daniel, Markus M. Schmid, and Heinz Zimmermann (2009) A Generalization of the
Calendar Time Portfolio Approach and the Performance of Private Investors, Unpublished Working Paper.
Hong, Harrison, Jeffrey D. Kubik, and Jeremy C. Stein (2008) The Only Game in Town –
Stock Price Consequences of Local Bias, Journal of Financial Economics 90 (1), 20–
37.
Huberman, Gur (2001) Familiarity Breeds Investment, Review of Financial Studies 14 (3),
659–680.
Ivkovic, Zoran and Scott Weisbenner (2005) Local Does As Local Is: Information Content of
the Geography of Individual Investors’ Common Stock Investments, Journal of Finance
60 (1), 267–306.
Jacobs, Heiko, and Martin Weber (2010) The Market Impact of Local Bias: Evidence form a
Natural Experiment, Unpublished Working Paper.
Jaffe, Jeffrey P. (1974) Special Information and Insider Trading, Journal of Business 47 (3),
410–428.
Karolyi, Andrew, and Rene M. Stulz (2003) Are Financial Assets Prices Locally or Globally?
Handbook of the Economics of Finance, G. Constantinides, M. Harris, and R. M. Stulz,
eds., Elsevier North Holland.
Korniotis, George M., and Alok Kumar (2009) Do portfolio distortions reflect superior information or psychological biases? Unpublished Working Paper.
Loughran, Tim, and Paul Schultz (2005) Liquidity: Urban versus Rural Firms, Journal of Financial Economics 78 (2), 341–374.
Lyon, John D., Brad M. Barber, and Chih-Ling Tsai (1999) Improved Methods for Tests of
Long-Run Abnormal Stock Returns, Journal of Finance 54 (1), 165–202.
Mandelker, Gershon N. (1974) Risk and Return: The Case of Merging Firms, Journal of Financial Economics 1 (4), 303–335.
Massa, Massimo, and Andrei Simonov (2006) Hedging, Familiarity, and Portfolio Choice,
Review of Financial Studies 19 (2), 633–685.
Odean, Terrance (1999) Do investors trade too much?, American Economic Review 89, 1279–
1298.
Rogers, William H. (1993) Regression Standard Errors in Clustered Samples, Stata Technical
Bulletin 13, 19–23.
Schwert, G. William (1990) Stock Volatility and the Crash of '87, Review of Financial Studies
3(1), 77–102.
Seasholes, Mark S., and Ning Zhu (2010) Individual Investors and Local Bias, Journal of
Finance 65 (5), 1987–2010.
26
Thomas, R. William, and Richard J. Hugett (1980) Modelling in Geography. A Mathematical
Approach, Barnes and Noble Books, Totowa, New Jersey.
Van Nieuwerburgh, Stijn, and Laura Veldkamp (2009) Information Immobility and the Home
Bias Puzzle, Journal of Finance 64 (3), 1187–1215.
Wengler, Martin (2006) Wechselwirkungen zwischen der finanzintermediatorischen Tätigkeit
kommunaler Kreditinstitute und dem kommunalen Raum, Special Issue, Halle Institute
for Economic Research.
Zhu, Ning (2002) The Local Bias of Individual Investors, ICF Working Paper No. 02-30,
Yale University.
27
Appendix
Calculation of returns employed in the performance analysis
As detailed in section 2.1, holdings data is reported at the end of each quarter. Following Coval and Moskowitz (2001), we thus update a given investor’s portfolio holdings
at the beginning of every quarter on the basis of the holdings reported for the previous
quarter, and assume them to remain unchanged over the subsequent three months. So,
for instance, the stock positions from the last quarter of 2008 (ending December 2008)
are used with return data for January, February, and March 2009. Specifically, the two
returns for investor i in quarter t are calculated as
Rlocal ,i ,t =  j ∈N i w iact, j ,t −1rj ,t
and
Rremote ,i ,t =  j ∉N i w iact, j ,t −1rj ,t
where R local ,i ,t and R remote ,i ,t are the returns over the quarter t on investor i’s local and
nonlocal stockholdings, respectively. N i reflects the number of stocks local to investor
i, w iact, j ,t −1 is the rescaled (to sum to one) fraction of stock j in her portfolio at the end of
quarter t-1, and, finally, r j ,t is the three-month raw return of stock j at time t. Each investor produces a time series of 17 quarters of local returns.
28
Figure 1
Geographical distribution of German individual investors and public limited companies
56
56
55
55
54
54
Latitude
Latitude
53
53
52
52
51
51
50
50
49
49
48
48
47
47
55
77
99
11
13
15
Longitude
Longitude
This figure plots the spatial coordinates (in degrees) of the private households (blue rhombuses)
as well as the public limited companies (red squares) represented in the sample. Households are
mapped within the zip code area of their custodian bank. Companies are mapped according to
the geographical location of their headquarters.
29
Figure 2
Changes in local bias among German individual investors during the sample period
0.8%
0.6%
Change local bias
0.4%
0.2%
0.0%
-0.2%
-0.4%
Active rebalancing
Passive rebalancing
-0.6%
Total change
-0.8%
2005Q4
2006Q4
2007Q4
2008Q4
2009Q4
This figure plots the evolution of average local bias levels across German private households for
the period between end-2005 and end-2009. Overall quarter-to-quarter changes (red line) are
dissected in trading-based shifts (active portfolio rebalancing) and shifts induced by price
movements absent transactions (passive portfolio rebalancing).
30
Figure 3
Local bias of German individual investors under changing market conditions
60
10.0%
VDAX New
9.5%
Local Bias
40
9.0%
30
8.5%
20
8.0%
10
7.5%
0
2005Q4
2006Q4
2007Q4
2008Q4
Local bias
VDAX New
50
7.0%
2009Q4
This figure plots German private households' quarterly local bias levels for the period between
end-2005 and end-2009 against the VDAX New index. the VDAX New captures implied return
volatility of the major German stock index DAX30 at a one-month horizon. Local bias levels are
adjusted for stock price movements between the reporting periods and thus are confined to
investors' active portfolio rebalancing decisions.
31
32
5,419,831
412,780
Western states
New Länder
1,066
649
1,570
145
Rural
Western states
New Länder
1,715
Urbanized
All
Geographical area of
investor residence
All
1,256,721
Rural
Panel B: Custodian banks
4,575,890
5,832,611
Total number
of portfolios
Urbanized
All
Geographical area of
investor residence
Panel A: Portfolio statistics
1,518
6,606
5,307
6,885
6,289
Median
59
374
140
293
433
Savings banks
Total number of banks
1,800
7,675
6,130
7,804
7,183
Mean
86
1,196
509
773
1,282
Cooperative banks
1,082 - 2,109
4,870 - 8,888
3,605 - 7,247
4,965 - 9,351
4,394 - 8,611
25th - 75th
Amount (EUR)
6.5
17.1
16.4
16.5
16.4
Median
4.6 - 9.5
12.2 - 24.0
11.0 - 23.1
11.2 - 23.4
11.1 - 23.3
25th - 75th
2,847
3,452
1,936
4,293
3,401
(continued on next page)
Average number household
portfolios per bank
8.4
19.8
18.5
19.0
18.8
Mean
Fraction of total portfolio value (%)
Domestic common stockholdings
Table 1
Summary statistics of sampled investor portfolios, custodian banks and companies
33
109
84
46
43
39
21
14
7
6
624
1,109
Munich
Berlin
Hamburg
Cologne
Düsseldorf
Stuttgart
Bremen
Dortmund
Essen
Other
Total
100.0
56.3
0.5
0.6
1.3
1.9
3.5
3.9
4.1
7.6
9.8
10.5
%
Company headquarters
100.0
43.7
43.2
42.6
41.3
39.4
35.9
32.0
27.9
20.3
10.5
Cum. %
This table presents basic characteristics of the private households and firms represented in the sample. Panel A reports descriptive portfolio statistics of
households delineated by the geographical area in which they reside, distinguishing between urban and rural, as well as West German states and East
German states ('New Länder'). Numbers in Panel A reflect averages across all years under review. Panel B presents summary statistics pertaining to the
custodian banks, to which the sampled households are affiliated. Finally, Panel C provides information on the geographic spread of the sampled public
limited companies headquarters within the ten largest cities in Germany as well as outside those areas (Other ).
116
N
Frankfurt
City
Panel C: Public limited companies
Table 1
Summary statistics of sampled investor portfolios, custodian banks, and companies―continued
Table 2
Locality of German individual investors' domestic stockholdings
Panel A: Distance between investor location and company headquarters (km)
Sample period
2005Q4 - 2009Q4
Average distance to
N
27,819
Holdings
Market portfolio
255.1
290.5
Difference
-35.4 ***
Panel B: Proportion of companies within investor locality sphere (%)
Average percentage of
Sample period
2005Q4 - 2009Q4
N
27,819
Holdings within 100
km
Market portfolio
within 100 km
20.1%
11.8%
Difference
8.3% ***
This table presents portfolio statistics of German private households referring to the locality of
their domestic stock investments. Panel A displays households' average distance to their actual
holdings and to the market portfolio built from the full stock universe under review (both distances
weighted by free float market value). The difference between the two distances is shown in the
rightmost column. Panel B reports the average fraction of a household's stockholdings which is
invested locally, i.e. within a radius of 100 kilometers around the household's place of residence,
as well as the proportional free-float market capitalization within the household's local range. The
difference between the two percentages denotes the local bias and is reported in the rightmost
column. *** indicates statistical significance at the 1%-level.
34
Table 3
Locality of German individual investors' domestic stockholdings, by investor location and
index status
Average percentage of
N
All
27,819
Holdings within
100 km
Market portfolio
within 100 km
20.1%
11.8%
Difference
8.3% ***
Panel A: Segmentation by geographical area in which private household resides
Urbanized
17,507
22.4%
12.7%
9.7% ***
Rural
10,312
16.3%
10.5%
5.8% ***
West German states
25,488
21.6%
12.9%
8.8% ***
2,331
3.8%
0.9%
2.9% ***
New Länder
Panel B: Segmentation by index status of stockholdings of private households' portfolio
DAX-listed
27,819
17.7%
12.4%
5.3% ***
Non-DAX-listed
27,819
25.1%
11.5%
13.7% ***
This table presents further details with regard to the locality of households' domestic stock
investments. The first row reports numbers for the overall sample: column two shows the number of
individual bank-quarter observations, the following columns report the average percentage of the
investor's actual local stockholdings, the portion of aggregate domestic market capitalization located
within 100 kilometers of the investor, and, finally, the local bias, denoted as the difference between
the two percentages (*** indicates statistical significance at the 1% level). Panel A reports the
corresponding figures for subsamples built according to the households' geographic area of
residence, distinguishing between urban and rural, as well as West German states and East German
states ('New Länder'). Panel B delineates households' stockholdings according to their index status,
differentiating shares listed in the German national stock index DAX30 and the remainder of stocks.
35
Table 4
Portfolio performance of German individual investors' local stockholdings
(Holdings-based portfolios, 3-month returns)
Regressions with Rlocal,i - R f as the dependent Variable
Reg 1
Reg 2
Reg 3
Reg 4
Panel A: All companies
Alpha (bp)
-1.30
(-0.41)
-0.24
(-0.28)
1.02 ***
(15.97)
BM
Rall
- Rf
BM
Rlocal,i
- Rf
Number of observations
Number of reporting quarters
27,819
17
27,819
17
-0.25
(-0.39)
0.89 ***
(28.41)
-0.14
(-0.24)
0.41 ***
(6.70)
0.62 ***
(10.18)
27,819
17
27,819
17
Panel B: Non-DAX companies
Alpha (bp)
-1.80
(-0.54)
-0.66
(-0.49)
0.97 ***
(9.57)
BM
Rall
- Rf
0.81 ***
(14.97)
BM
Rlocal,i
- Rf
Number of observations
Number of reporting quarters
-0.92
(-0.85)
27,819
17
27,819
17
27,819
17
-0.67
(-0.63)
0.55 ***
(5.87)
0.45 ***
(5.56)
27,819
17
This table reports pooled regression results (clustered by quarters) of the analysis of German private
households' equity portfolio performance , with the excess return on the local portion of a
household's equity portfolio (Rlocal,i - R f) as the dependent variable. In equations Reg 2 and Reg 3,
BM
this excess return is regressed on an overall market return ( Rlocal,i
- R f ) and an investor-specific local
BM
benchmark (R - R ), respectively. Reg 4 represents the full specification including both benchmark
all
f
types. T-statistics are based on Rogers (1993) standard errors and are robust to heteroskedasticity
and contemporaneous correlation. Returns are adjusted for dividend payouts and stock splits and
companies under review are not subject to survivorship bias. Panel A reports results for the full
universe of sampled domestic stocks, while results in panel B exclude the 30 largest German publicly
listed companies (members of the major national stock index DAX30). *** indicates statistical
significance at the 1% level.
36
Table 5
Portfolio performance for investor quartiles formed on local bias levels
(Holdings-based portfolios, 3-month returns)
Quartile 1
(Low)
Local bias
Alpha (bp)
BM
Rall
- Rf
BM
Rlocal,i
- Rf
Number of
observations
Number of
reporting quarters
2
R
-0.9%
Quartile 2
Quartile 3
3.1%
8.0%
Quartile 4
(High)
High - Low
21.5%
22.4% ***
-0.59
(-0.93)
0.5143 ***
(6.02)
0.4836 ***
(6.77)
-0.34
(-0.58)
0.3972 ***
(8.16)
0.6296 ***
(9.58)
-0.30
(-0.05)
0.2901 ***
(5.40)
0.7108 ***
(9.85)
0.11
(0.22)
0.2185 ***
(4.07)
0.8629 ***
(13.99)
6,955
6,955
6,955
6,954
17
17
17
17
0.56
0.69
0.77
0.80
0.70
-0.2958 ***
0.3793 ***
This table assigns the results of Regression 4 of Table 4 to quartiles according to German individual
investors' local bias, where Quartile 1 (Quartile 4) subsumes the least (most) locally biased private
households in the sample. The upper row reports mean local bias levels for each of the four
portfolios. Regression results are displayed in the following rows. Differences between Quartile 1 and
Quartile 4 are presented in the rightmost column. *** indicates statistical significance at the 1% level.
37
Table 6
Portfolio performance of German individual investors' local stockholdings
(Transaction-based portfolios, 12-month returns)
Dependent variable:
buy
local,i
R
sell
local,i
Dependent variable:
buy
sell
Rremote,i
- Rremote,i
-R
Reg 1
Reg 2
-5.75 **
(-2.45)
-5.26 **
(-2.62)
0.06
(1.01)
Reg 1
Reg 2
Panel A: All companies
Alpha (bp)
BM
Rlocal,i
- Rf
-3.13
(-0.41)
0.26
(1.54)
BM
Rremote,i
- Rf
Number of observations
Number of reporting quarters
-0.70
(-0.09)
20,112
16
20,112
16
20,112
16
20,112
16
Panel B: Non-DAX companies
Alpha (bp)
-5.57 ***
(-4.00)
-4.78 ***
(-4.21)
0.09 **
(2.64)
BM
Rlocal,i
- Rf
-8.90 ***
(-4.45)
0.10 **
(2.23)
BM
Rremote,i
- Rf
Number of observations
Number of reporting quarters
-7.85 ***
(-4.30)
18,542
16
18,542
16
18,542
16
18,542
16
This table reports pooled regression results (clustered by quarter) of the analysis of the performance
of portfolios constructed from private households' aggregate buys and sells, using 12-month returns.
The dependent variable in columns 2 and 3 expresses the difference between returns on the
buy
sell
household's local buy- and sell-positions ( Rlocal,i
). The dependent variable in columns 4 and 5
- Rlocal,i
expresses the difference between returns on the household's nonlocal buy- and sell-positions
buy
sell
( Rremote,i
). Panel A reports results for the full universe of sampled domestic stocks, while
- Rremote,i
results in panel B exclude the 30 largest German publicly listed companies (members of the major
national stock index DAX30). Returns are adjusted for dividend payouts and stock splits and
companies under review are not subject to survivorship bias. *** indicates statistical significance at
the 1% level.
38
Table 7
Impact of stock market uncertainty on local bias among German individual investors
Regressions with quarter-to-quarter percentage change of
local bias as dependent variable
change local bias (t ? 1)
change VDAX New (t)
Reg 1
Reg 2
0.327 *
(2.01)
0.057 ***
(3.05)
0.159
(0.60)
0.065 ***
(3.38)
0.030
(0.98)
change VDAX New (t ? 1)
Number of observations
2
Adjusted R
16
0.31
16
0.30
This table reports regression results of the impact of stock market uncertainty (as captured by
quarterly changes of the average 3-month-VDAX New levels, change VDAX New ) on changes in
local bias levels of German individual investors' common stockholdings. T-statistics are based on
Newey-West standard errors. *** indicates statistical significance at the 1% level.
39
Table 8
Market performance and local bias among German individual investors
Average change of
CDAX
Average change of
local bias
Quartile 1
-16.32%
2.75%
Quartile 2
-2.94%
2.43%
Quartile 3
6.78%
1.09%
Quartile 4
14.39%
-2.27%
This table classifies the period under review into four quartiles according to the average stock
market performance, as captured by the quarterly return of the Composite DAX index (CDAX).
40
Table 9
Variation in local bias changes across German individual investors
Total
number of
banks
Number of banks with
changing local bias levels, by
direction of change
Increase
Decrease
Ratio of
increased local
bias
Change of
local bias
2006Q1
1,705
898
802
52.7%
-0.7%
2006Q2
1,698
1,141
552
67.2%
4.3%
2006Q3
1,679
1,014
660
60.4%
2.2%
2006Q4
1,669
966
699
57.9%
0.4%
2007Q1
1,663
904
757
54.4%
1.6%
2007Q2
1,661
980
681
59.0%
1.6%
2007Q3
1,645
860
784
52.3%
1.0%
2007Q4
1,636
923
713
56.4%
1.1%
2008Q1
1,636
915
721
55.9%
1.3%
2008Q2
1,629
1,030
599
63.2%
3.4%
2008Q3
1,610
1,068
542
66.3%
3.8%
2008Q4
1,594
919
675
57.7%
6.4%
2009Q1
1,592
734
857
46.1%
-0.2%
2009Q2
1,581
703
878
44.5%
-2.0%
2009Q3
1,562
444
1,118
28.4%
-7.6%
2009Q4
1,548
740
808
47.8%
0.1%
This table dissects private households in the sample according to whether they increase or
reduce their exposure to local stocks between two consecutive quarters of the period under
review (end-2005 through end-2009). Households' stockholdings are aggregated at the bank
level. The rightmost column reports the percentage change in average local bias levels across all
investors.
41
The following Discussion Papers have been published since 2010:
Series 1: Economic Studies
01
2010
Optimal monetary policy in a small open
economy with financial frictions
Rossana Merola
02
2010
Price, wage and employment response
to shocks: evidence from the WDN survey
Bertola, Dabusinskas
Hoeberichts, Izquierdo, Kwapil
Montornès, Radowski
03
2010
Exports versus FDI revisited:
Does finance matter?
C. M. Buch, I. Kesternich
A. Lipponer, M. Schnitzer
04
2010
Heterogeneity in money holdings
across euro area countries:
the role of housing
Ralph Setzer
Paul van den Noord
Guntram Wolff
05
2010
Loan supply in Germany
during the financial crises
U. Busch
M. Scharnagl, J. Scheithauer
06
2010
Empirical simultaneous confidence
regions for path-forecasts
Òscar Jordà, Malte Knüppel
Massimiliano Marcellino
07
2010
Monetary policy, housing booms
and financial (im)balances
Sandra Eickmeier
Boris Hofmann
08
2010
On the nonlinear influence of
Reserve Bank of Australia
interventions on exchange rates
Stefan Reitz
Jan C. Ruelke
Mark P. Taylor
09
2010
Banking and sovereign risk
in the euro area
S. Gerlach
A. Schulz, G. B. Wolff
10
2010
Trend and cycle features in German
residential investment before and after
reunification
Thomas A. Knetsch
42
11
2010
What can EMU countries’ sovereign
bond spreads tell us about market
perceptions of default probabilities
during the recent financial crisis?
Niko Dötz
Christoph Fischer
Tobias Dümmler
Stephan Kienle
12
2010
User costs of housing when households face
a credit constraint – evidence for Germany
13
2010
Extraordinary measures in extraordinary times –
public measures in support of the financial
Stéphanie Marie Stolz
sector in the EU and the United States
Michael Wedow
14
2010
The discontinuous integration of Western
Europe’s heterogeneous market for
corporate control from 1995 to 2007
Rainer Frey
15
2010
Bubbles and incentives:
Ulf von Kalckreuth
a post-mortem of the Neuer Markt in Germany Leonid Silbermann
16
2010
Rapid demographic change and the allocation
of public education resources: evidence from
East Germany
Gerhard Kempkes
17
2010
The determinants of cross-border bank flows
to emerging markets – new empirical evidence Sabine Herrmann
on the spread of financial crisis
Dubravko Mihaljek
18
2010
Government expenditures and unemployment: Eric Mayer, Stéphane Moyen
a DSGE perspective
Nikolai Stähler
19
2010
NAIRU estimates for Germany: new evidence
on the inflation-unemployment trade-off
Florian Kajuth
20
2010
Macroeconomic factors and
micro-level bank risk
43
Claudia M. Buch
Sandra Eickmeier, Esteban Prieto
21
22
2010
2010
How useful is the carry-over effect
for short-term economic forecasting?
Karl-Heinz Tödter
Deep habits and the macroeconomic effects
of government debt
Rym Aloui
23
2010
Price-level targeting
when there is price-level drift
C. Gerberding
R. Gerke, F. Hammermann
24
2010
The home bias in equities
and distribution costs
P. Harms
M. Hoffmann, C. Ortseifer
25
2010
Instability and indeterminacy in
a simple search and matching model
Michael Krause
Thomas Lubik
26
2010
Toward a Taylor rule for fiscal policy
M. Kliem, A. Kriwoluzky
27
2010
Forecast uncertainty and the
Bank of England interest rate decisions
Guido Schultefrankenfeld
01
2011
Long-run growth expectations
and “global imbalances”
M. Hoffmann
M. Krause, T. Laubach
02
2011
Robust monetary policy in a
New Keynesian model with imperfect
interest rate pass-through
Rafael Gerke
Felix Hammermann
The impact of fiscal policy on
economic activity over the business cycle –
evidence from a threshold VAR analysis
Anja Baum
Gerrit B. Koester
Classical time-varying FAVAR models –
estimation, forecasting and structural analysis
S. Eickmeier
W. Lemke, M. Marcellino
03
04
2011
2011
44
05
2011
The changing international transmission of
financial shocks: evidence from a classical
time-varying FAVAR
Sandra Eickmeier
Wolfgang Lemke
Massimiliano Marcellino
06
2011
FiMod – a DSGE model for
fiscal policy simulations
Nikolai Stähler
Carlos Thomas
07
2011
Portfolio holdings in the euro area –
home bias and the role of international,
domestic and sector-specific factors
Axel Jochem
Ute Volz
08
2011
Seasonality in house prices
F. Kajuth, T. Schmidt
09
2011
The third pillar in Europe:
institutional factors and individual decisions
Julia Le Blanc
10
2011
In search for yield? Survey-based
evidence on bank risk taking
C. M. Buch
S. Eickmeier, E. Prieto
11
2011
Fatigue in payment diaries –
empirical evidence from Germany
Tobias Schmidt
Christoph Fischer
12
2011
Currency blocs in the 21st century
13
2011
How informative are central bank assessments Malte Knüppel
of macroeconomic risks?
Guido Schultefrankenfeld
14
2011
Evaluating macroeconomic risk forecasts
Malte Knüppel
Guido Schultefrankenfeld
15
2011
Crises, rescues, and policy transmission
through international banks
Claudia M. Buch
Cathérine Tahmee Koch
Michael Koetter
16
2011
Substitution between net and gross settlement
systems – A concern for financial stability?
Ben Craig
Falko Fecht
45
17
18
2011
2011
Recent developments in quantitative models
of sovereign default
Nikolai Stähler
Exchange rate dynamics, expectations,
and monetary policy
Qianying Chen
19
2011
An information economics perspective
on main bank relationships and firm R&D
D. Hoewer
T. Schmidt, W. Sofka
20
2011
Foreign demand for euro banknotes
issued in Germany: estimation using
direct approaches
Nikolaus Bartzsch
Gerhard Rösl
Franz Seitz
21
2011
Foreign demand for euro banknotes
issued in Germany: estimation using
indirect approaches
Nikolaus Bartzsch
Gerhard Rösl
Franz Seitz
22
2011
Using cash to monitor liquidity –
implications for payments, currency
demand and withdrawal behavior
Ulf von Kalckreuth
Tobias Schmidt
Helmut Stix
23
2011
Home-field advantage or a matter of
ambiguity aversion? Local bias among
German individual investors
Markus Baltzer
Oscar Stolper
Andreas Walter
46
Series 2: Banking and Financial Studies
01
2010
Deriving the term structure of banking
crisis risk with a compound option
approach: the case of Kazakhstan
Stefan Eichler
Alexander Karmann
Dominik Maltritz
02
2010
Recovery determinants of distressed banks:
Regulators, market discipline,
or the environment?
Thomas Kick
Michael Koetter
Tigran Poghosyan
03
2010
Purchase and redemption decisions of mutual
fund investors and the role of fund families
Stephan Jank
Michael Wedow
04
2010
What drives portfolio investments of
German banks in emerging capital markets?
Christian Wildmann
05
2010
Bank liquidity creation and
risk taking during distress
Berger, Bouwman
Kick, Schaeck
06
2010
Performance and regulatory effects of
non-compliant loans in German synthetic
mortgage-backed securities transactions
Gaby Trinkaus
Banks’ exposure to interest rate risk, their
earnings from term transformation, and
the dynamics of the term structure
Christoph Memmel
Completeness, interconnectedness and
distribution of interbank exposures –
a parameterized analysis of the stability
of financial networks
Angelika Sachs
Do banks benefit from internationalization?
Revisiting the market power-risk nexus
C. M. Buch
C. Tahmee Koch, M. Koetter
07
08
09
2010
2010
2010
47
10
2010
Do specialization benefits outweigh
concentration risks in credit portfolios
of German banks?
Rolf Böve
Klaus Düllmann
Andreas Pfingsten
11
2010
Are there disadvantaged clienteles
in mutual funds?
Stephan Jank
12
2010
Interbank tiering and money center banks
Ben Craig, Goetz von Peter
13
2010
Are banks using hidden reserves
to beat earnings benchmarks?
Evidence from Germany
Sven Bornemann, Thomas Kick
Christoph Memmel
Andreas Pfingsten
14
2010
How correlated are changes in banks’ net
interest income and in their present value?
Christoph Memmel
Contingent capital to strengthen the private
safety net for financial institutions:
Cocos to the rescue?
George M. von Furstenberg
01
2011
02
2011
Gauging the impact of a low-interest rate
environment on German life insurers
Anke Kablau
Michael Wedow
03
2011
Do capital buffers mitigate volatility
of bank lending? A simulation study
Frank Heid
Ulrich Krüger
04
2011
The price impact of lending relationships
Ingrid Stein
05
2011
Does modeling framework matter?
A comparative study of structural
and reduced-form models
Yalin Gündüz
Marliese Uhrig-Homburg
Contagion at the interbank market
with stochastic LGD
Christoph Memmel
Angelika Sachs, Ingrid Stein
06
2011
48
07
2011
The two-sided effect of financial
globalization on output volatility
Barbara Meller
08
2011
Systemic risk contributions:
a credit portfolio approach
Klaus Düllmann
Natalia Puzanova
09
2011
The importance of qualitative risk
assessment in banking supervision
before and during the crisis
Thomas Kick
Andreas Pfingsten
10
2011
Bank bailouts, interventions, and
moral hazard
Lammertjan Dam
Michael Koetter
11
2011
Improvements in rating models
for the German corporate sector
Till Förstemann
The effect of the interbank network
structure on contagion and common shocks
Co-Pierre Georg
12
2011
49
Visiting researcher at the Deutsche Bundesbank
The Deutsche Bundesbank in Frankfurt is looking for a visiting researcher. Among others
under certain conditions visiting researchers have access to a wide range of data in the
Bundesbank. They include micro data on firms and banks not available in the public.
Visitors should prepare a research project during their stay at the Bundesbank. Candidates
must hold a PhD and be engaged in the field of either macroeconomics and monetary
economics, financial markets or international economics. Proposed research projects
should be from these fields. The visiting term will be from 3 to 6 months. Salary is
commensurate with experience.
Applicants are requested to send a CV, copies of recent papers, letters of reference and a
proposal for a research project to:
Deutsche Bundesbank
Personalabteilung
Wilhelm-Epstein-Str. 14
60431 Frankfurt
GERMANY
50